Knowledge Graph–Enhanced Deep Learning for Investigating Covid-19 Pathogenesis
by Deepthi Rani S. S., Dr. Renu Aggarwal
Published: May 4, 2026 • DOI: 10.51584/IJRIAS.2026.110400061
Abstract
Knowledge graphs (KGs), which represent entities and their relationships in a structured semantic network, have been widely applied in the study of various diseases such as thyroid disorders, cardiovascular diseases, and neurological conditions. However, current diagnostic approaches often face challenges including incomplete data integration, limited scalability, and reduced diagnostic accuracy. These limitations highlight the need for advanced methodologies capable of addressing the complex nature of COVID-19 diagnosis.The proposed research introduces a framework that integrates knowledge graphs with deep learning techniques for improved COVID-19 analysis. Initially, COVID-19 related datasets will be collected from publicly available repositories such as Kaggle, including information on viral characteristics, transmission patterns, clinical symptoms, and public health data. From these datasets, relevant entities and relationships will be extracted to construct a COVID-19-specific knowledge graph. The constructed KG will then be transformed into low-dimensional vector representations using embedding techniques, enabling semantic representation within a vector space.Based on this knowledge representation, a novel deep learning model will be developed to predict COVID-19 cases using relevant input features. The model will utilize virus-related word vectors and knowledge entity vectors derived from the knowledge graph. Through supervised learning, the model will be trained to classify samples based on COVID-19 related symptoms and associated features. The effectiveness of the proposed diagnostic model will be evaluated using standard performance metrics.By integrating knowledge graph construction with deep learning models, the proposed study aims to improve understanding of COVID-19 pathogenesis and support evidence-based decision making in pandemic management. This approach has the potential to provide an efficient and accurate diagnostic tool for the early detection and management of COVID-19 cases.